On Tue, 29 Mar 2005 22:48:12 +0100, Stefek Zaba
wrote:

Lee wrote:

The one we had prior used the Rugby signal, and that was better at
getting a signal, but worse at updating...

My bugbear with these is the pretty analogue ones - one at work
intermittently, and one at home consistently, is off by one minute. You
take out the battery and reinsert it, the hands sweep majestically round
to the nearest of 12, 4, or 8 o'clock, and sit waiting for sync. A
coupla minutes later, sync it does: but while the seconds are perfect,
the minute hand is one position back from where it should be.

It seems like a fault in either the way the minute hand's been put on
the shaft (but surely there's some sort of positioning key!?) or the
shaft encoder or whatever feedback mechanism it uses (surely there's
*some* sort of feedback mech to get a reference position for the
hands!?), since when I say 'sweep majestically round to ... o'clock',
the consistently-faulty one actually sweeps them round majestically to
one minute *before* the Starting Position. I've not dared try to pull
the hands right off and reposition (well, OK, I *did* dare try, but
didn't dare pull more than very gently ;-), and I keep this off-by-one
clock in my study, where I know how to interpret its quirk ;-) But it's
the sort of thing which sets my perfectionist teeth on edge!

Stefek


They probably run WindowsCE......

Since there are now more and more devices around with embedded 802.11,
I am surprised that nobody has thought of doing a clock which syncs by
periodically doing an NTP request.



--

..andy

To email, substitute .nospam with .gl